Khaled Koutini

Khaled Koutini is a Post-Doctoral Researcher at the Institute of Computational Perception, Johannes Kepler University Linz. His research focuses on learning general-purpose audio representations using deep neural networks, addressing challenges like data scarcity and improving generalization through inductive biases in architectures and training processes. He explores transfer learning applications in audio classification/tagging for multimedia, smart devices, and monitoring systems.

His current PhD project (ongoing since 2020), supervised by Gerhard Widmer, investigates enhancing audio recognition via inductive biases and representation extraction from large-scale models. Affiliated with the Institute of Computational Perception, his work bridges theoretical advancements in neural networks with practical applications in machine listening.